Abstract. Support Vector Machines (SVM) have been applied successfully in a wide variety of fields in the last decade. The SVM problem is formulated as a convex objective function...
In multiple-instance learning (MIL), an individual example is called an instance and a bag contains a single or multiple instances. The class labels available in the training set ...
We empirically study the relationship between supervised and multiple instance (MI) learning. Algorithms to learn various concepts have been adapted to the MI representation. Howe...
We present a new approach to multiple instance learning (MIL) that is particularly effective when the positive bags are sparse (i.e. contain few positive instances). Unlike other ...
Abstract. The aim of this paper is to present a new tool of multiple instance learning which is designed using a grammar based genetic programming (GGP) algorithm. We study its app...